fight or flight flight increases immune gene expression but does not help to fight an infection

immune gene expression in response to 15-min forced flight treatment impacts individual survival of bacterial infection in the Glanville fritillary butterfly Melitaea cinxia.. We were ab

Fight or flight? – Flight increases immune gene expression but does not help to fight an infection

L W O E S T M A N N * , J K V I S T† & M SAASTAMOINEN*

*Metapopulation Research Centre, University of Helsinki, Helsinki, Finland

†Institute of Biotechnology, University of Helsinki, Helsinki, Finland

Keywords:

gene expression;

immune response;

insect flight;

Melitaea cinxia.

Abstract

Flight represents a key trait in most insects, being energetically extremely demanding, yet often necessary for foraging and reproduction Additionally, dispersal via flight is especially important for species living in fragmented landscapes Even though, based on life-history theory, a negative relation-ship may be expected between flight and immunity, a number of previous studies have indicated flight to induce an increased immune response In this study, we assessed whether induced immunity (i.e immune gene expression) in response to 15-min forced flight treatment impacts individual survival of bacterial infection in the Glanville fritillary butterfly (Melitaea cinxia) We were able to confirm previous findings of flight-induced immune gene expression, but still observed substantially stronger effects on both gene expression levels and life span due to bacterial infection compared to flight treatment Even though gene expression levels of some immunity-related genes were elevated due to flight, these individuals did not show increased survival of bacterial infection, indicating that flight-induced immune activation does not completely protect them from the negative effects of bacterial infection Finally, an interaction between flight and immune treatment indicated a potential trade-off: flight treatment increased immune gene expression in na€ıve individuals only, whereas in infected indi-viduals no increase in immune gene expression was induced by flight Our results suggest that the up-regulation of immune genes upon flight is based

on a general stress response rather than reflecting an adaptive response to cope with potential infections during flight or in new habitats

Introduction

Parasites and pathogens represent a strong selection

pressure to the host, as they are ubiquitous and can

cause substantial fitness costs (Decaestecker et al., 2007;

Mone et al., 2010) Therefore, the evolution of the

immune system is a crucial factor in the life of any

spe-cies The investment of an organism in its immune

defence depends on several factors such as the risk of

an attack and the efficiency of the defence but also on

the costs associated with the activation of the immune

system (Zuk & Stoehr, 2002) Further, immunity investment might be affected by individual’s body condition or nutritional status (Klemola et al., 2007; Valtonen et al., 2009; Srygley & Lorch, 2011)

Another key life-history trait in many organisms is dispersal, playing a major role in population dynamics,

as it is a prerequisite for spreading of individuals and hence of gene flow among populations (Clobert et al., 2012) Dispersal includes several functions, such as escape from unfavourable conditions or habitats, avoid-ance of kin competition or inbreeding, but it also dis-tributes offspring into different locations and different environmental conditions (Matthysen, 2012) In many insects, flight is a key prerequisite for dispersal

As both flight and activation of immunity are ener-getically demanding, potential trade-offs between them may be expected (Bonte et al., 2012) Studies with

Correspondence: Luisa Woestmann, Metapopulation Research Centre,

University of Helsinki, PO Box 65, Viikinkaari 1, 00014 Helsinki,

Finland.

Tel.: +3580504484423;

e-mail: luisa.woestmann@helsinki.fi

ª 2 0 1 6 T H E A U T H O R S J E V O L B I O L 3 0 ( 2 0 1 7 ) 5 0 1 – 5 1 1

501

J O U R N A L O F E V O L U T I O N A R Y B I O L O G Y P U B L I S H E D B Y J O H N W I L E Y & S O N S L T D O N B E H A L F O F E U R O P E A N S O C I E T Y F O R E V O L U T I O N A R Y B I O L O G Y

T H I S I S A N O P E N A C C E S S A R T I C L E U N D E R T H E T E R M S O F T H E C R E A T I V E C O M M O N S A T T R I B U T I O N - N O N C O M M E R C I A L - N O D E R I V S L I C E N S E , W H I C H

crickets (Gryllus texensis) and bumblebees (Bombus

ter-restris), for example, have shown reduced immune

defence after foraging or tethered flight (K€oning &

Schmid-Hempel, 1995; Adamo et al., 2008), potentially

due to energetic costs of flight However, positive

corre-lations between flight and immunity have also been

observed (Snoeijs et al., 2004; Suhonen et al., 2010)

For example, in the great tit (Parus major) immigrants

have higher humoral immune response (Snoeijs et al.,

2004) The positive relationship between immunity and

flight may be an adaptive response allowing individuals

to cope with a potentially increasing infection risk due

to dispersal, for example, because entering a new

habi-tat may entail different quality or quantity of

patho-gens In such case, flight-induced immune activation

should increase individual’s survival to pathogens

Alternatively, up-regulation of immunity genes may

solely reflect a general stress response due to the

weari-some and stressful act of flight

In insects, the immune system is triggered by surface

particles of pathogens that are able to bind to secreted

but also membrane-bound receptors that can be found

in the haemolymph (Yoshida et al., 1996) Two main

pathways are part of the insect immune system, the

Toll and the IMD pathway, of which the latter responds

to gram-negative and the former to gram-positive

bac-teria and fungi (reviewed in Lemaitre & Hoffmann,

2007) Upon receptor binding, the Sp€atzle protein gets

activated via a proteolytic cascade which then binds to

the Toll receptor on the cell surface Contrarily,

anti-gens are able to bind directly to cell surface receptors in

the IMD pathway to then transmit the signal inside the

cell In both cases, the intracellular signalling cascade

leads to the activation of transcription factors (Dorsal

and Relish for Toll and IMD, respectively) that alter

gene expression of different immune genes (Hoffmann,

2003) Different proteins and molecules will be

expressed in the fat body and secreted into the

haemo-lymph, for example antimicrobial peptides (AMPs) and

serpins

A general stress response has previously been shown

to interact with the immune response in many insect

species (Adamo, 2008, 2012) This connection has

either evolved independently in different phyla or

rep-resents a conserved connection (Adamo, 2008) and

seems to be crucial for survival in many species The

NF-kB system, one of the key regulators of the innate

immune system, for example, is closely connected to

oxidative stress and inflammation (Salminen et al.,

2008) During acute stress (fight-or-flight), different

stress hormones are released, of which in insects the

most important are octopamine and adipokinetic

hor-mone (Orchard et al., 1993) Both horhor-mones trigger the

release of lipids from the fat body to optimize the body

for a fight-or-flight reaction Immune challenge

like-wise leads to the increase in octopamine to increase the

availability of energy-rich compounds (Adamo, 2010)

The liberation of lipids might result in a shift of molec-ular resources away from immunity into the fight-or-flight response, as both pathways rely on apolipophorin III, a lipoprotein that is responsible for lipid transport (Adamo et al., 2008) This protein has both storage and immune function, as it is able to bind bacterial lipotei-choic acid (Kim et al., 2004; Ma et al., 2006) Stress hormones tend to also increase immune responses such

as increased phagocytosis and phenoloxidase response (Baines et al., 1992; Goldsworthy et al., 2002), most likely due to stress-hormone receptors on haemocytes (Adamo, 2008; Kim et al., 2009; Huang et al., 2012) In larvae of the greater wax moth (Galleria mellonella), acute stress had an immune-enhancing effect even

24 h after a 2-min stress event (Mowlds et al., 2008)

In this study, we aimed to disentangle why individu-als would invest in an up-regulation of the immune system upon flight, and more specifically whether the activation is based on a general stress response or on an adaptive response that may have evolved along with a higher infection risk when dispersing to new environ-ments We are using the Glanville fritillary (Melitaea cinxia) as a study system, which has a classic metapopu-lation structure in the Aland Islands in the south-west

of Finland The metapopulation is characterized by annual extinctions and recolonizations of local popula-tions, making dispersal essential for population viability

in a highly fragmented landscape (Hanski, 1999a) Flight in this species is energetically demanding and might impact the individual’s condition, therefore plac-ing dispersplac-ing individuals at a higher risk of infections Dispersing individuals may also be facing different or more infections by parasites and pathogens in the habi-tat matrix or in the new habihabi-tat patches they disperse

to The energetic demands of flight might further increase pathogen exposure by increased food uptake after flight events Previous studies in this species have shown that forced flight provokes an activation of the immune system, measured as higher encapsulation rate (Saastamoinen & Rantala, 2013) In addition, immune genes are up-regulated upon forced flight treatment (Kvist et al., 2015)

We infected adult butterflies with a bacterial strain right after a forced flight treatment to try to tease apart whether induced immunity upon flight mitigates indi-viduals to overcome infection, and hence show similar

or longer life span than those without flight treatment

As an immune response, we assessed gene expression

of seven immune genes that have been previously shown to be expressed upon forced flight treatment similar to that used in the present experiment (Kvist

et al., 2015), and which are known to be expressed upon infection with bacteria A similar pattern to that

in life span should be visible in the gene expression if the adaptive response hypothesis is true, hence show-ing equally high or even higher expression levels for the flight treatment in comparison with control

individuals when facing an infection The alternative

hypothesis is that flight-induced immune activation is

due to a general stress response, in which case we

expect no positive effect on survival or immune gene

expression due to flight, and might in fact expect a

neg-ative effect if the stress is severe enough We were

aim-ing to cover a wide range of immune genes, includaim-ing

recognition proteins, antimicrobial peptides, and

recep-tors and proteins involved in both the Toll and IMD

pathway to see whether the observed responses were

general or pathway- or gene-specific Whereas the

receptor PGRP-LC and the AMP attacin are part of the

IMD pathway, pelle is involved in the Toll pathway

We furthermore included lysozyme, prophenoloxidase

(proPO), serpin andb-1,3-glucan recognition protein in

our study, based on these criteria

Materials and methods

Study system

The Glanville fritillary butterfly, Melitaea cinxia

(Meli-taeini: Nymphalidae), is present in Finland only on the



Aland Islands south-west of mainland Finland where it

has a classic metapopulation (Hanski, 1999b) The

spe-cies is characterized by a univoltine life cycle in Finland

Larvae feed for five instars on one of two host plants

(Plantago lanceolata and Veronica spicata) before they

spend the winter in diapause in a silken web In the

spring, larvae continue feeding until pupation occurs in

May following by a flight season from June to mid-July,

with males emerging about 2–3 days earlier than females

(Boggs & Nieminen, 2004) Adults feed on nectar

In general, this species is relatively sedentary (based

on mark–release–recapture studies), although many

individuals move between the small meadows at some

point during the adult stage (Kuussaari et al., 1996)

Flight typically consists of short flight bouts and rapid

take-offs in case of males that locate females by the

‘perching’ tactic, in which they establish a mating

terri-tory waiting for a suitable female that they defend from

intruding males by chasing them away In addition,

males might fly more continuously around the habitat

in search of females (Boggs & Nieminen, 2004) The

average flight distance experimentally measured was

about 32 m and mean lifetime dispersal distance several

hundreds of metres with longest dispersal events of

1–2 km (Kuussaari et al., 1996; Niitep~old et al., 2011)

and the longest recorded colonization distance of 4–

5 km (Van Nouhuys & Hanski, 2002) Females show

higher rates of dispersal compared to males that might

remain in the natal population (Kuussaari et al., 1996)

Experimental design

Larvae were collected from 58 different populations (one

to five individuals per population, on average 2.07) in

 Aland in the spring 2015 and fed ad libitum with P lance-olata until pupation Upon eclosion, 386 adult butterflies from different populations (families) were randomly divided into a flight treatment and control group, with

an equal sex distribution Individuals were kept in cages (409 50 cm) with no more than 40 individuals per cage and fed on 20% honey:water solution Butterflies were kept at room temperature to discourage flight activity and food uptake on the second day of eclosion to stan-dardize nutritional state as well as activity On the third day after eclosion, butterflies in the flight treatment were placed into a cylindrical plastic chamber and allowed to acclimatize to the chamber before the treatment Individ-uals were forced to fly actively for 15 min by gently tap-ping or shaking the chamber whenever the butterfly landed The temperature during flight was maintained at

30°C This treatment reflects the general flight metabolic rate measurement assay often used in this species (Niitep~old et al., 2009) The individuals not included in the flight treatment were not flown but otherwise trea-ted equally (i.e placed in the chamber with the same temperature) The 3-day-old adults from both control and flight treatment groups were then randomly divided across three different immunity treatments: na€ıve, injec-tion of 2lL PBS into the thorax (wounding control) or injection of 2lL of a 5-mg mL 1

lyophilized Micrococcus luteus (ATCC No 4698; Sigma-Aldrich) solution into the thorax The butterfly was spanned with a net on a soft sponge with ventral side up to ensure that it is not able

to move Through a small hole in the net, the thorax is accessible for the injection with a Hamilton syringe Here

as well, na€ıve individuals were placed on the sponge under the net, even though not injected After the differ-ent immune treatmdiffer-ents, butterflies were provided with 20% honey:water solution and kept in standardized con-ditions, avoiding dehydration or other stress that might influence gene expression Individuals from the three different immunity treatment groups were randomly divided into (1) measurement of life span or (2) assessing immune gene expression and therefore RNA sampling Individuals whose life span was assessed were provided with 20% honey:water solution, and survival was checked daily

RNA sampling Twenty hours after the flight treatment, individuals that had been randomly chosen for RNA sampling were killed by flash-freezing them in liquid nitrogen Individ-uals across all treatments (flight and control and the three immune treatments) have been used for RNA sampling Based on the 20-h incubation, samples were taken between 8 and 12 am Control individuals that did not experience flight treatment were similarly sam-pled 20 h after placing them once into the flight cham-ber All samples were stored at 80°C until RNA extraction from the thorax

RNA extraction and reverse transcription

Total RNA was extracted from the frozen thorax using

TRIzol reagent (Life Technologies Corporation,

Carls-bad, CA, USA) followed by extraction with acid

–phe-nol:chloroform:isoamyl alcohol (24 : 24 : 1, pH 5) and

chloroform Precipitation of the RNA was performed

using isopropanol, washed with 75% ethanol, air-dried

and resuspended in 35–50 lL MQ water RNA quantity

and quality were checked using NanoDrop (Thermo

Fischer Scientific Inc., Waltham, MA, USA) Samples

were stored at 80°C until further usage Potential

contaminations of genomic DNA in the RNA samples

were removed using DNase I (Thermo Fischer Scientific

Inc.) The samples were then reverse-transcribed to

cDNA using iScriptTM

cDNA Synthesis Kit (Bio-Rad Lab-oratories, Hercules, CA, USA) according to the

manu-facturer’s instructions

Quantitative real-time PCR (qPCR)

Primers were designed with Primer3 (Rozen & Skaletsky,

2000) for seven immune response genes: lysozyme C

(MCINX003391), prophenoloxidase (proPO; MCINX013403),

attacin (MCINX009397), peptidoglycan recognition

pro-tein LC (PGRP-LC; MCINX014869),b-1,3-glucan

recog-nition protein (bGRP; MCINX012854), serpin 3a

(MCINX005220) and pelle (MCINX001775); and three

endogenous control genes: mitochondrial ribosomal

protein L37 (MCINX003184) and S24 (MCINX003139)

and histone variant H2A.Z (MCINX016093) All

pri-mers were ordered from Oligomer (Oligomer Oy,

Hel-sinki, Finland) The sequences can be found in the

supporting information (Appendix S1) Amplification

efficiencies (E) of the primer pairs were determined

with five dilutions (1 : 1, 1 : 5, 1 : 25, 1 : 125, 1 : 625)

of template cDNA, where E= 10-1/slope The qPCR

was performed with three technical replicates, one

water control and a plate control sample in a 384-well

plate with 10lL volume, using C1000TM Thermal

Cycler (Bio-Rad Laboratories) All samples were tested

for genomic DNA contamination with -RT controls

prior to qPCR Each reaction used 1lL of the 1 : 5

diluted cDNA, 5lL of SYBRGreen containing master

mix (iQTM

SYBR Green Supermix for qPCR; Bio-Rad

Laboratories), 3lL of nuclease-free water and 0.5 lL of

each primer (10lM)

Statistical analysis

Immune gene expression for each sample was

calcu-lated relative to the geometric mean of the three

refer-ence genes For each sample, the mean from the three

technical replicates was used, with the exception of

removing a possible outlier Raw Ct values for all genes

and technical replicates can be found in the supporting

information (Appendix S2) A linear mixed-model

approach (R 3.1.2 for Windows; The R Project for Sta-tistical Computing; lmer from package lme4; Bates et al., 2015) was used to analyse the effects of flight treatment and infection on immune gene expression, using bacte-rial treatment, flight treatment, sex and gene as fixed factors and family (population) as a random term In a subsequent analysis (due to three-way interaction between flight treatment, bacterial treatment and gene)

to explore the effect for every gene separately, an inde-pendent model for each of the immune genes was con-ducted with bacterial treatment, flight treatment and sex as fixed factors and family (population) as a ran-dom term Post hoc analysis was performed to explore paired comparisons of the different treatment groups The model with the lowest Akaike information criterion (AIC) value was chosen as the best fitting model, and the model fit was further assessed using the conditional

R2(sem.model.fits from package piecewiseSEM; Lefcheck

& Freckleton, 2016) AIC values and R2 of the final models are shown in Table 1 and in Appendix S3 for the initial full models

The effect of flight or infection on the life span was analysed using Poisson distribution with glmmPQL to handle overdispersion (package MASS; Venables & Rip-ley, 2002), using bacterial and flight treatment and sex

as fixed factors and family (population) as a random term Backward model selection was used by starting with a full model including all meaningful second-order interactions and sequentially eliminating nonsignificant interaction terms (P> 0.05) that did not improve the model

Results

Immune gene expression

We found a significant three-way interaction between flight treatment, bacterial treatment and gene (v2

Appendix S4) that was further explored with a gene-by-gene analysis A significant increase in four of the seven immune genes in the bacteria-exposed groups relative to the na€ıve groups was observed (P < 0.003 for all; Fig 1; Table 1) A strong up-regulation was detected for attacin, showing on average 540-fold increase (log2FC = 9.08) in expression in the bacteria-exposed group compared to na€ıve individuals Pelle and bGRP likewise showed a strong up-regulation with on average 22- to 26-fold increase (log2FC= 4.65 & 4.48)

in expression in the bacteria-exposed group compared

to na€ıve individuals A moderate up-regulation was detected for serpin with on average 2.5-fold increase (log2FC= 1.29) in expression due to bacterial injection Wounding itself led to an increase in expression levels for attacin and pelle only (P< 0.02 for both; Fig 1; Table 1), with on average 85- and two-fold increase (log2FC= 6.41 and 1.22) compared to the na€ıve group,

respectively In addition, we found a significant

interac-tion between bacterial treatment and sex for bGRP

(bacterial treatment*sex: v2

Appendix S5), showing higher expression levels for the

bacterial treatment for females Finally, flight treatment

provoked an increase in expression levels forbGRP and

PGRP-LC (bGRP: v2

v2

the expression of attacin and pelle in the na€ıve samples,

whereas in the infected samples no such elevation was

observed (bacterial treatment*flight treatment: attacin:

v2

P = 0.0003; Appendix S5) If anything, in the infected

individuals the expression was slightly reduced by the

flight treatment

There were no significant changes for lysozyme and proPO All results of the initial models used for the gene-by-gene analysis can be found in Appendix S3

Life span Males lived longer than females (t1,171= 2.11,

P= 0.037; Fig 2) Life span was significantly reduced

in both sexes by the bacterial infection treatment (by almost 66%, P< 0.00001; Fig 2) compared to na€ıve individuals (na€ıve: 24.9 ( 1.6) and 23.7 ( 1.6); bac-teria: 11.9 ( 1.5) and 8.3 ( 1.4); life span in days for males and females, respectively) Injection of PBS had

no significant effect on life span in either sex (P> 0.1; males: 23.1 ( 1.7) and females: 19.4 ( 1.5) days),

Table 1 Relative expression levels and fold increase for the used immune genes divided by the different treatment groups.

Relative expression

AIC = 414.58; R 2

AIC = 419.61; R 2

AIC = 421.55; R 2

AIC = 423.07; R 2

AIC = 312.23.07; R 2

Expression levels are calibrated to na €ıve individuals without flight treatment, and sexes are pooled Significant effects are highlighted in bold and calculated with Post hoc analysis (Tukey honest significant differences).

−5

0

5

10

15

20

−5

0

5

10

15

20

−5

0

5

10

15

20

−5

0

5

10

15

20

(g) PGRP-LC

(d) Serpin (c) proPO

*

Flight Injection

P = 0.0007 P<0.0001

Flight Injection

P = 0.45

Flight Injection

P = 0.23

P = 0.001

Flight Injection Injection*Flight

P = 0.006

P = 0.0003

Flight Injection

P = 0.63

Flight Injection Injection*Flight

P = 0.79 P<0.0001

P = 0.0006

Flight Injection

P = 0.04

P = 0.15

Fig 1 Immune gene expression in the Glanville fritillary butterfly Shown are the relative expression levels (log 2 -transformed) of the seven tested immune genes of na €ıve (white), PBS-injected (blue; wounding with 2 lL PBS) and bacteria-injected (red; 2 lL of 5-mg mL 1Micrococcus luteus in PBS) individuals divided into control (C) and flight (F) treatment groups Expression levels are calibrated to the na €ıve individuals that did not experience flight treatment Sexes are combined, as

no sex difference was observed The P-values for the effect of flight and immune treatment and their interaction, whenever significant, are presented The interactions between flight and immune treatment found for attacin and pelle are indicated with asterisks and lines based on the Post hoc test performed.

showing that even though infection was not performed

under sterile conditions, the wounding itself did not

significantly affect individual life span Flight treatment

did not influence life span (t1,171= 1.12, P = 0.26;

Fig 2)

Discussion

In this study, we aimed to disentangle whether the

commonly observed positive relationship between

flight and immune response also previously found in

our study species (Saastamoinen & Rantala, 2013; Kvist

et al., 2015) is a by-product of a general stress response

or potentially reflects an adaptive response that

evolved along with higher infection risk due to

disper-sal events We found a significant increase in

expres-sion levels in four of the seven tested immune genes

as well as reduced life span based on bacterial

infec-tion Flight itself led to an increase in expression levels

of some of the immune genes; however, it had no

influence on individual survival upon infection

Fur-thermore, the up-regulation of immune genes after

flight treatment was far lower than that induced by

infection with bacteria or wounding Our findings

therefore suggest that the increased immune gene

expression upon flight is most likely due to a general

stress response rather than an adaptive response to

possible upcoming infections

As we are unaware of any specific natural pathogen

affecting the Glanville fritillary butterfly, we used a

simple and rather general bacterial strain in our study

The used strain successfully infects our species and

leads to reduced survival and increased immune gene expression in both sexes Most strains of Micrococcus luteus are gram-positive (Madigan et al., 2015) and hence should trigger the Toll pathway The increased expression of pelle upon bacterial injection supports the activation of this pathway As b-1,3-glucan is an anti-gen on the surface of gram-positive bacteria, the increased expression levels for b-1,3-glucan recognition protein are likewise expected for this bacterial strain (Jiang et al 2004) Also, peptidoglycan recognition pro-teins (PGRP) are known to be strongly up-regulated upon bacterial infection in the fruit fly (Drosophila mela-nogaster; Zaidman-Remy et al., 2011) and in the silk-worm (Bombyx mori; Tanaka et al., 2008) In contrast,

we did not observe a significant increase in expression

of PGRP-LC upon bacterial infection PGRP-LC is a receptor in the IMD pathway and therefore theoreti-cally not affected by gram-positive bacteria However, this receptor detects peptidoglycan of bacteria, and gen-erally, both pathways do interact in case of an infection (Lemaitre & Hoffmann, 2007) The investigated time frame of the expression levels may have been too short

to detect a significant up-regulation of PGRP-LC Lyso-zyme is an enLyso-zyme damaging the cell wall of both gram-positive and gram-negative bacteria and poten-tially activates the Toll pathway by the release of com-ponents of the bacterial cell wall (Dunn, 1986; Hultmark, 2003) We did not detect changes in expres-sion of lysozyme activity induced by our treatments There are several lysozyme genes in M cinxia as in most organisms of which some are more specific to the bac-terial cell wall Potentially, the chosen gene in this

10

Time (d)

Naive C PBS C Bacteria C

Naive F PBS F Bacteria F

(b) Males

0

20

40

60

80

100

Time (d)

Naive C PBS C Bacteria C

Naive F PBS F Bacteria F

(a) Females

Fig 2 Survival of the adult (a) females and (b) males in days Solid lines indicate individuals without flight treatment (C), whereas dashed lines indicate those with flight treatment (F) prior to injection Na €ıve group is presented in black (n = 17 (C) and n = 19 (F) for females and n = 25 (C) and n = 24 (F) for males) in comparison with the group injected with 2 lL PBS in blue (n = 16 (C) and n = 17 (F) for females and n = 22 (C) and n = 24 (F) for males) and the group injected with 2 lL of a 5-mg mL 1

Micrococcus luteus solution in PBS in red (n = 18 (C) and n = 19 (F) for females and n = 24 (C) and n = 23 (F) for males) Only bacterial injection resulted in a significant reduction

of the life span for both sexes.

study was not strongly affected by our immune

treat-ment In general, up-regulation of lysozyme should last

long as it has been even detected 120 h after bacterial

infection in the mosquito Anopheles gambiae (Kajla et al

2010), suggesting that the timing is not the issue

Prophenoloxidase represents the main regulator of the

melanization pathway in insects (Cerenius & S€oderhall,

2004) As for lysozyme activity, no changes in

expres-sion of proPO upon bacterial infection after 20 h were

observed As intermediate products of the melanin

syn-thesis are toxic (Cerenius & S€oderhall, 2004),

melaniza-tion is a tightly regulated process in insects The

pathway needs to be suppressed quickly to avoid

possi-ble damage to own tissue such as cuticles and wings or

in severe cases even the death of the individual (De

Gregorio et al., 2002) Certain inhibitors exist to

pre-vent extended expression of the melanization pathway

Serine protease inhibitors, also known as serpins,

repre-sent such molecules and are able to bind to the

prophe-noloxidase-activating enzyme (PPAE) that normally

converts inactive prophenoloxidase to the active form

phenyloxidase (De Gregorio et al., 2002) The here

tested serpin gene was significantly up-regulated upon

bacterial infection Together with the observation that

proPO did not show any increase in expression, we

hypothesize that increased levels of serpin have led to

the inactivation of PPAE to avoid potential tissue

dam-age Thus, an earlier time point likely would show an

increase in proPO due to bacterial infection and

poten-tially a slightly lower level of serpin

The flight treatment itself resulted in higher

expres-sion levels for two of the tested immune genes,bGRP

and PGRP-LC Somewhat surprisingly, expression of

these genes was not elevated by flight in the previous

study of the same species (Kvist et al., 2015) Consistent

with the previous study, no increase in expression due

to flight was found 20 h after the flight treatment for

proPO and lysozyme Most of the used genes in this

study are known to be strongly up-regulated due to

infection However, flight might not trigger such

expression levels, as it reflects cellular stress and is

therefore less specific Interestingly, we found a

signifi-cant interaction between flight and bacterial treatment

influencing the expression levels of attacin and pelle

For both of these genes, flight did induce higher

expression levels but only in na€ıve individuals In

indi-viduals that were infected with bacteria, flight either

caused no effect or even reduced the expression levels

This slight decrease in expression of attacin and pelle

due to the flight treatment in infected individuals might

indicate a resource trade-off, and the infected

individu-als simply cannot further induce their immune gene

expression when flying Similarly, a trade-off between

flight and immunity has been observed, for example, in

the migratory monarch butterfly (Danaus plexippus),

where such a decrease in number of haemocytes was

found in individuals that experienced tethered flight assay compared to a control group (Fritzsche McKay

et al., 2016) The increase in the expression levels for attacin and pelle due to the flight treatment in the

na€ıve individuals is in accordance with the previous study (Kvist et al., 2015) It was suggested that forced flight potentially shifted molecular resources (such as apolipophorin III) away from immune response towards flight-related functions in case of an infection (Adamo

et al., 2008) Notably, however, in the present study the expression level induced by flight in combination with infection is still higher than that of flight treat-ment on the na€ıve individual We therefore hypothe-size that the elevated immune gene expression upon flight in the Glanville fritillary butterfly is acting in maintaining the maximal immune defence while at the same time optimizing the individuals for a fight-or-flight response

We only used one bacterial strain in the present study, and potentially, the results may be different if a different strain, for example more specific to Lepi-doptera, would have been used Similarly, injection of bacteria into the haemocoel is not a natural way of infection in the wild However, butterflies do experi-ence injuries, for example due to predators, which might allow pathogens to enter directly into the haemocoel and provoke a faster and more drastic immune response Our injection treatment therefore mimics such an event in nature To ensure that our effect was not just provoked by wounding alone, we included a wounding control in the experimental

set-up Wounding led to an increased expression of two immune genes that were likewise triggered by the bac-terial treatment (see Fig 1), as indicated by the PBS treatment However, the up-regulation was substan-tially higher with bacterial infection compared to wounding only, suggesting that introduction of bacteria definitely provoked a stronger immune response and led to an infection The additional life span data con-firm that wounding itself might activate the immune response but does not significantly reduce life span, which occurred upon bacterial infection Finally, the flight treatment used in this study does not reflect a natural dispersal event; however, experimentally induced flight can produce important knowledge on the costs and benefits of flight (Chapman et al., 2015)

As gene expression levels change readily, we treated individuals equally in terms of handling, acclimation and feeding across all treatments

Although we do not provide a complete picture of the effects of infection and flight on immunity, our study provides interesting new insights in trying to understand why different organisms in some cases up-regulate their immune system upon or during flight or dispersal Even though previous studies have shown that immunity is often up-regulated during flight, these

studies have not looked at whether that up-regulation

actually benefits individuals in case of infection We

found no indication that the up-regulation of immunity

genes due to flight would help individuals to recover or

survive from infection On the contrary, we detect a

potential trade-off between flight and immunity for

two genes, attacin and pelle Applying our findings to

other systems, especially those with known natural

pathogens, would be highly interesting Here, the

mon-arch butterfly (D plexippus) and the protozoan

Ophry-ocystis elektroscirrha are interesting candidates Further

studies in the Glanville fritillary butterfly, especially

within the metapopulation framework, would be

use-ful For example, comparing the relationship between

dispersal and immunity among individuals from old

and newly established populations or from continuous

vs fragmented landscapes that are known to differ in

dispersal ability (Saastamoinen, 2007; Somervuo et al.,

2014) would be highly relevant Similarly, further

experiments investigating the immune response in

combination with disease models assessing

epidemiol-ogy in this metapopulation could bring new insights

into disease dynamics in the wild

Ethics statement

The Glanville fritillary butterfly is not classified as an

endangered or protected species No permits are

required for the collection of individuals in the Aland

Islands

Acknowledgment

We acknowledge Suvi Ikonen and Kati Schenk for

their help with the experiment, Mikko Frilander for

helpful discussions, Kristjan Niitep~old for comments

on the manuscript and Toshka Nyman for RNA

extractions

Funding

This study was funded by grants from the European

Research Council (Independent Starting grant

META-STRESS; 637412) and the Academy of Finland

(Deci-sion numbers 273098 and 265641) to MS

Competing interest

The authors declare that they have no competing

interests

Authors’ contribution

LW, JK and MS designed the study; LW performed the

experiment; LW analysed the data; and LW, JK and MS

wrote the manuscript

References Adamo, S.A 2008 Norepinephrine and octopamine: linking stress and immune function across phyla Invertebrate Surviv.

J 5: 12 –19.

Adamo, S.A 2010 Why should an immune response activate the stress response? Insights from the insects (the cricket Gryllus texensis) Brain Behav Immun 24: 194 –200.

Adamo, S.A 2012 The effects of the stress response on immune function in invertebrates: an evolutionary perspec-tive on an ancient connection Horm Behav 62: 324 –330 Adamo, S.A., Roberts, J.L., Easy, R.H & Ross, N.W 2008 Competition between immune function and lipid transport for the protein apolipophorin III leads to stress-induced immunosuppression in crickets J Exp Biol 211: 531 –538 Baines, D., DeSantis, T & Downer, R.G.H 1992 Octopamine and 5-hydroxytryptamine enhance the phagocytic and nod-ule formation activities of cockroach (Periplaneta americana) haemocytes J Insect Physiol 38: 905 –914.

Bates, D., Maechler, M., Bolker, B.M & Walker, S 2015 Fit-ting linear mixed-effects models using lme4 J Stat Softw arXiv:1406.5823

Boggs, C.L & Nieminen, M 2004 Checkerspot Reproductive Biol-ogy Oxford University Press, Oxford.

Bonte, D., Van Dyck, H., Bullock, J.M., Coulon, A., Delgado, M., Gibbs, M., et al 2012 Costs of dispersal Biol Rev Camb Philos Soc 87: 290 –312.

Cerenius, L & S€oderhall, K 2004 The prophenoloxidase-acti-vating system in invertebrates Immunol Rev 198: 116 –126 Chapman, J.W., Reynolds, D.R & Wilson, K 2015 Long-range seasonal migration in insects: mechanisms, evolutionary dri-vers and ecological consequences Ecol Lett 18: 287 –302 Clobert, J., Baguette, M., Benton, T.G & Bullock, J.M 2012 Dispersal Ecology and Evolution Oxford University Press, Oxford, UK.

De Gregorio, E., Han, S.J., Lee, W.-J., Baek, M.-J., Osaki, T., Kawabata, S.-I., et al 2002 An immune-responsive serpin regulates the melanization cascade in drosophila Dev Cell 3:

581 –592.

Decaestecker, E., Gaba, S., Raeymaekers, J.A., Stoks, R., Van Kerckhoven, L., Ebert, D & De Meester, L 2007 Host-para-site ‘Red Queen’ dynamics archived in pond sediment Nat-ure 450: 870 –873.

Dunn, P.E 1986 Biochemical aspects of insect immunity Annu Rev Entomol 31: 321 –339.

Fritzsche McKay, A., Ezenwa, V.O & Altizer, S 2016 Unrav-elling the costs of flight for immune defenses in the migra-tory monarch butterfly Integr Comp Biol 56: 278 –289 Goldsworthy, G., Opoku-Ware, K & Mullen, L 2002 Adipoki-netic hormone enhances laminarin and bacterial lipopolysac-charide-induced activation of the prophenoloxidase cascade

in the African migratory locust, Locusta migratoria J Insect Physiol 48: 601 –608.

Hanski, I 1999a Habitat connectivity, habitat continuity, and metapopulations in dynamic landscapes Oikos 87:

209 –219.

Hanski, I 1999b Metapopulation Ecology Oxford University Press, New York.

Hoffmann, J.A 2003 The immune response of Drosophila Nature 426: 33 –38.

Huang, J., Wu, S.F., Li, X.H., Adamo, S.A & Ye, G.Y 2012 The characterization of a concentration-sensitive

alpha-adrenergic-like octopamine receptor found on insect

immune cells and its possible role in mediating stress

hor-mone effects on immune function Brain Behav Immun 26:

942 –950.

Hultmark, D 2003 Drosophila immunity: paths and patterns.

Curr Opin Immunol 15: 12 –19.

Jiang, H., Ma, C., Lu, Z.-Q & Kanost, M.R 2004

b-1,3-Glu-can recognition protein-2 ( bGRP-2) from Manduca sexta: an

acute-phase protein that binds b-1,3-glucan and

lipotei-choic acid to aggregate fungi and bacteria and stimulate

prophenoloxidase activation Insect Biochem Mol Biol 34:

89 –100.

Kajla, M.K., Andreeva, O., Gilbreath, T.M & Paskewitz, S.M.

2010 Characterization of expression, activity and role in

antibacterial immunity of Anopheles gambiae lysozyme c-1.

Comp Biochem Physiol B Biochem Mol Biol 155: 201 –209.

Kim, G.S., Nalini, M., Kim, Y & Lee, D.W 2009 Octopamine

and 5-hydroxytryptamine mediate hemocytic phagocytosis

and nodule formation via eicosanoids in the beet armyworm,

Spodoptera exigua Arch Insect Biochem Physiol 70: 162 –176.

Kim, H.J., Je, H.J., Park, S.Y., Lee, I.H., Jin, B.R., Yun, H.K.,

et al 2004 Immune activation of apolipophorin-III and its

distribution in hemocyte from Hyphantria cunea Insect

Bio-chem Mol Biol 34: 1011 –1023.

Klemola, N., Klemola, T., Rantala, M.J & Ruuhola, T 2007.

Natural host-plant quality affects immune defence of an

insect herbivore Entomol Exp Appl 123: 167 –176.

Kuussaari, M., Nieminen, M & Hanski, I 1996 An

experi-mental study of migration in the glanville fritillary butterfly

Melitaea cinxia J Anim Ecol 65: 791 –801.

Kvist, J., Mattila, A.L., Somervuo, P., Ahola, V., Koskinen, P.,

Paulin, L., et al 2015 Flight-induced changes in gene

expression in the Glanville fritillary butterfly Mol Ecol 24:

4886 –4900.

K€oning, C & Schmid-Hempel, P 1995 Foraging activity and

immunocompetence in workers of the bumble bee, Bombus

terrestris L Proc Biol Sci 260: 225 –227.

Lefcheck, J.S & Freckleton, R 2016 piecewiseSEM: piecewise

structural equation modelling in r for ecology, evolution,

and systematics Methods Ecol Evol 7: 573 –579.

Lemaitre, B & Hoffmann, J 2007 The host defense of

Droso-phila melanogaster Annu Rev Immunol 25: 697 –743.

Ma, G., Hay, D., Li, D., Asgari, S & Schmidt, O 2006

Recog-nition and inactivation of LPS by lipophorin particles Dev.

Comp Immunol 30: 619 –626.

Madigan, M.T., Martinko, J.M., Bender, K.S., Buckley, D.H &

Stahl, D.A 2015 Brock Biology of Microorganisms, 14th edn.

Pearson, Boston.

Matthysen, E 2012 Multicausality of Dispersal: A Review Oxford

University Press, Oxford, UK.

Mone, Y., Gourbal, B., Duval, D., Du Pasquier, L.,

Kieffer-Jaquinod, S & Mitta, G 2010 A large repertoire of parasite

epitopes matched by a large repertoire of host immune

receptors in an invertebrate host/parasite model PLoS Negl.

Trop Dis 4: e813 doi: 10.1371/journal.pntd.0000813.

Mowlds, P., Barron, A & Kavanagh, K 2008 Physical stress

primes the immune response of Galleria mellonella larvae to

infection by Candida albicans Microb Infect 10: 628 –634.

Niitep ~old, K., Mattila, A.L.K., Harrison, P.J & Hanski, I 2011.

Flight metabolic rate has contrasting effects on dispersal in

the two sexes of the Glanville fritillary butterfly Oecologia 165: 847 –854.

Niitep ~old, K., Smith, A.D., Osborne, J.L., Reynolds, D.R., Car-reck, N.L., Martin, A.P., et al 2009 Flight metabolic rate and Pgi genotype influence butterfly dispersal rate in the field Ecology 90: 2223 –2232.

Orchard, I., Ramirez, J.-M & Lange, A.-B 1993 A multifunc-tional role for octopamine in locust flight Annu Rev Ento-mol 38: 227 –249.

Rozen, S & Skaletsky, H 2000 Primer3 on the WWW for general users and for biologist programmers Methods Mol Biol 132: 365 –386.

Saastamoinen, M 2007 Mobility and lifetime fecundity in new versus old populations of the Glanville fritillary butterfly Oecologia 153: 569 –578.

Saastamoinen, M & Rantala, M.J 2013 Influence of develop-mental conditions on immune function and dispersal-related traits in the Glanville fritillary (Melitaea cinxia) butterfly PLoS One 8: e81289.

Salminen, A., Huuskonen, J., Ojala, J., Kauppinen, A., Kaarni-ranta, K & Suuronen, T 2008 Activation of innate immu-nity system during aging: NF-kB signaling is the molecular culprit of inflamm-aging Ageing Res Rev 7: 83 –105 Snoeijs, T., Van de Casteele, T., Adriaensen, F., Matthysen, E.

& Eens, M 2004 A strong association between immune responsiveness and natal dispersal in a songbird Proc Biol Sci 271(Suppl 4): S199 –S201.

Somervuo, P., Kvist, J., Ikonen, S., Auvinen, P., Paulin, L., Koskinen, P., et al 2014 Transcriptome analysis reveals sig-nature of adaptation to landscape fragmentation PLoS One 9: e101467 doi: 10.1371/journal.pone.0101467.

Srygley, R.B & Lorch, P.D 2011 Weakness in the band: nutrient-mediated trade-offs between migration and immu-nity of Mormon crickets, Anabrus simplex Anim Behav 81:

395 –400.

Suhonen, J., Honkavaara, J & Rantala, M.J 2010 Activation

of the immune system promotes insect dispersal in the wild Oecologia 162: 541 –547.

Tanaka, H., Ishibashi, J., Fujita, K., Nakajima, Y., Sagisaka, A., Tomimoto, K., et al 2008 A genome-wide analysis of genes and gene families involved in innate immunity of Bombyx mori Insect Biochem Mol Biol 38: 1087 –1110.

Valtonen, T.M., Kleino, A., R €amet, M & Rantala, M.J 2009 Starvation reveals maintenance cost of humoral immunity Evol Biol 37: 49 –57.

Van Nouhuys, S & Hanski, I 2002 Colonization rates and dis-tances of a host butterfly and two specific parasitoids in a fragmented landscape J Anim Ecol 71: 639 –650.

Venables, W.N & Ripley, B.D 2002 Modern Applied Statistics with S, 4th edn Springer, New York.

Yoshida, H., Kinoshita, K & Ashida, M 1996 Purification of

a peptidoglycan recognition protein from haemolymph from the silkworm Bombyx mori J Biol Chem 271:

13854 –13860.

Zaidman-Remy, A., Poidevin, M., Herve, M., Welchman, D.P., Paredes, J.C., Fahlander, C., et al 2011 Drosophila immu-nity: analysis of PGRP-SB1 expression, enzymatic activity and function PLoS One 6: e17231.

Zuk, M & Stoehr, A.M 2002 Immune defense and host life history Am Nat 160: S9 –S22.